This study is part of a program of research intended to evaluate and improve the degree to which physician training promotes physicians' ability to provide equally high quality and patient-centered care to all patients regardless of their race or ethnicity, size, or sexual orientation. Physician behavior and decision-making is a key contributor to the extensively documented racial disparities in health care. Despite evidence that physicians are motivated to provide high quality care to all of their patients, implicit (unconscious) biases may still be expressed subtly and indirectly, for example in physicians' assumptions about patient characteristics that affect treatment recommendations. These implicit biases, because they are typically unconscious, can have a systematic but largely unintentional impact on treatment and care. Many medical school and residency training programs have implemented curricula in hopes of preventing racial and other biases in trainees but approaches vary widely and have received little evaluation. Furthermore, little is known about how formal training interacts with informal norms and values. This study will address these gaps by assessing the impact of residency characteristics, independent of individual student and medical school factors, on implicit and explicit bias in a large longitudinal cohort of 3650 graduating medical students. This renewal builds on rich multi-measure data collected through repeated assessment of medical students enrolled in the Medical Student CHANGES Study (1 R01 HL085631, Contributors to Racial Bias in Medical Student Judgment and Decision-Making) by following the cohort through residency training. In CHANGES we found that training in racial bias, negative role-modelling, interracial contact, and diversity climate all predicted change in implicit racial bias between the 1st and 4th years of medical school. The degree to which these effects persist beyond medical school is unknown. We also found that medical students held high levels of negative implicit bias towards obese and sexual minority individuals that persisted independently of and in contrast to their conscious (explicit) attitudes. This study uses a longitudinal repeated measures design to assess the impact of residency characteristics, independently and in combination with individual and medical school factors, on new physicians': 1) racial, obesity and sexual orientation biases and 2) attitudes and behaviors that have been shown to protect from, or exacerbate, the impact of bias on behavior and decision-making. To our knowledge, this is the first longitudinal study of this scope that investigates changes in physician trainees' implicit biases and related cognitive and behavioral factors that can directly influence the quality of medical care they will provide. Understanding the factors that shape new physicians' social beliefs and attitudes across their medical training, particularly in the final stages in which they prepare for practice, will provide unique information for new interventions that can promote more effective health care for historically stigmatized groups.